Angular velocities models are necessary to analyse inter- and intra-deformations due to the secular motions of tectonic plates. They permit quantify the magnitude and direction of relative displacement between different tectonic units. The use of inaccurate models can produce incorrect conclusions, particularly at plate boundaries.

In the last decades, many angular velocity models have been produced based on geological/geophysical and/or geodetic measurements. The first ones (e.g., NUVEL-1A and MORVEL) average the motions of tectonic models in the last 3My whereas the geodetic ones (e.g., REVEL and GEODVEL) are representative of the present-day tectonic motions.

Due to the used methodology and data, geologic/geophysical models are only able to be estimated by inverting data for pairs of (or multiple) tectonic plates. This implies that the estimation is always relative to a reference plate. This is not the case for the geodetic models where the used data (normally velocities derived from space-based techniques, particularly GNSS observations) are estimated with respect to a No-Net-Rotation reference frame. Consequently, both approaches are possible: to estimate the angular motion of each plate individually or by inverting a global dataset of velocities.

In this work, we discuss the differences of both approaches by looking at the South America tectonic plate, which is the slowest one of the large tectonic plates. We compare our own dedicated estimations of the plate angular velocities for these plates with the predicted motions given by global plate models and other recent published models. We show that the differences can still be significant when different plate models are considered despite of the small magnitude of the South America, which can lead to incorrect interpretations, particularly on the deforming boundaries like in Colombia.


  • Fernandes, Rui (*); Bos, Machiel (*) Mora-Páez, Héctor (**); Corchuelo, Yuli (**); Giraldo-Londoño, Leidy (**); Gutierrez, Nancy (*); Audemard, Franck (***)